In current Tor design, everytime a user type in an URL1, TBB will connect with that URL1 using a Tor path (say, A-B1-C1, with C1 is an exit node). And, for another URL2, TBB will connect with that URL2 using a Tor path A-B2-C2, wich C2 is another exit node.

So my question is that, if the website of URL1 has contents that link to website of URL2 (e.g., a video clip, or a picture), will TBB still use C1 to load those contents or will it use C2 in background to load the contents from URL2?


Not quite.

It uses the "first party domain" to isolate requests (I.E. requests within a tab will use the same circuit, based on the domain in the tab).

If URL1 has some <img src=URL2> html entity, it should use C1 to fetch it, similarly if URL2 has some <img src=URL1> content, it should use C2 to fetch it.

Note that this doesn't guarantee different exits, just distinct circuits.

For more information see:

#3455 (Tor Browser should set SOCKS username for a request based on first party domain)

Cross-Origin Identifier Unlinkability

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.